Getting a new air filter is a lot like when you get a new pair of shoes as a kid. Despite there being almost no evidence to support it, you could swear that you’re going faster.

But it turns out, in the case of air filters, you are actually going faster, but only a little.

Jason Fenske of Engineering Explained got a dyno, four air filters, and a Subaru Crosstrek together and ran each filter to figure out if they were adding any power.

To no one’s surprise (but perhaps their relief), the K&N filter actually did make a reasonable amount of difference, adding more than 6 hp over the dirty filter. Not an enormous improvement, but not a rounding error, either.

Slightly more surprising (though, the youth of the “dirty” filter may help explain this), the clean OEM filter basically did nothing to improve hp, adding fewer than two horsepower.

The most surprising, though, was that the bargain option added 5 hp to the tally, only being beaten out slightly by the more expensive K&N option. Granted, Fenske’s tests didn’t account for longevity or filtration, it does seem to be the best hp bang for the buck.

Fenske explains that filters can add horspower in two ways. The first is simply by allowing more air through into the engine, which is a hot rodding trick that’s as old as time.

The second reason is related but has to do with the car finding more air and correcting the air/fuel ratio by adding more fuel, which in turn adds more horsepower.

So there you have it, better filters behave about as you’d expect, so it's not all in your imagination.